Identification of Two Mutations in Human Xanthine Dehydrogenase Gene Responsible for Classical Type I Xanthinuria

Identification of two mutations in human xanthine dehydrogenase gene responsible for classical type I xanthinuria.

K Ichida, … , T Hosoya, O Sakai

J Clin Invest. 1997;99(10):2391-2397. https://doi.org/10.1172/JCI119421.

Research Article

Hereditary xanthinuria is classified into three categories. Classical xanthinuria type I lacks only xanthine dehydrogenase activity, while type II and molybdenum cofactor deficiency also lack one or two additional enzyme activities. In the present study, we examined four individuals with classical xanthinuria to discover the cause of the enzyme deficiency at the molecular level. One subject had a C to T base substitution at nucleotide 682 that should cause a CGA (Arg) to TGA (Ter) nonsense substitution at codon 228. The duodenal mucosa from the subject had no xanthine dehydrogenase protein while the mRNA level was not reduced. The two subjects who were siblings with type I xanthinuria were homozygous concerning this mutation, while another subject was found to contain the same mutation in a heterozygous state. The last subject who was also with type I xanthinuria had a deletion of C at nucleotide 2567 in cDNA that should generate a termination codon from nucleotide 2783. This subject was homozygous for the mutation and the level of mRNA in the duodenal mucosa from the subject was not reduced. Thus, in three subjects with type I xanthinuria, the primary genetic defects were confirmed to be in the xanthine dehydrogenase gene.

Find the latest version: https://jci.me/119421/pdf

Identiﬁcation of Two Mutations in Human Xanthine Dehydrogenase Gene Responsible for Classical Type I Xanthinuria

Kimiyoshi Ichida,* Yoshihiro Amaya,‡ Naoyuki Kamatani,§ Takeshi Nishino,ʈ Tatsuo Hosoya,* and Osamu Sakai* *Second Department of Medicine, The Jikei University School of Medicine, Tokyo 105; ‡Department of Biochemistry, Yokohama City University School of Medicine, Yokohama 236, Japan; §Institute of Rheumatology, Tokyo Women’s Medical College, Tokyo 162; and ʈDepartment of Biochemistry and Molecular Biology, Nippon Medical School, Tokyo 113, Japan

Abstract enzyme itself is a typical molybdenum containing flavo-protein. Hereditary xanthinuria is classified into three categories. Inherited xanthine dehydrogenase deficiency, or xanthin- Classical xanthinuria type I lacks only xanthine dehydroge- uria, was first reported by Dent and Philport (1). The current nase activity, while type II and molybdenum cofactor defi- classification of this inherited disorder is rather complicated. ciency also lack one or two additional enzyme activities. In In classical xanthinuria type I, only xanthine dehydrogenase the present study, we examined four individuals with classi- activity is lacking, while in classical xanthinuria type II, alde- cal xanthinuria to discover the cause of the enzyme defi- hyde oxidase activity is also deficient (2, 3). More complicated ciency at the molecular level. One subject had a C to T base still is the presence of molybdenum cofactor deficiency, in substitution at nucleotide 682 that should cause a CGA which sulfite oxidase activity is missing as well as the above (Arg) to TGA (Ter) nonsense substitution at codon 228. The two enzymes (4). duodenal mucosa from the subject had no xanthine dehy- Classical xanthinuria types I and II are rare autosomal re- drogenase protein while the mRNA level was not reduced. cessive disorders and the combined incidence has been re- The two subjects who were siblings with type I xanthinuria ported to be 1/69,000 (5). The affected individuals may de- were homozygous concerning this mutation, while another velop urinary tract calculi, acute renal failure, or myositis due subject was found to contain the same mutation in a het- to tissue deposition of xanthine, but some subjects with ho- erozygous state. The last subject who was also with type I mozygous xanthinuria remain asymptomatic (2). Molybdenum xanthinuria had a deletion of C at nucleotide 2567 in cDNA cofactor deficiency is usually associated with severe neurologi- that should generate a termination codon from nucleotide cal disorders (4). The relationship between the three condi- 2783. This subject was homozygous for the mutation and tions has not been well understood at the molecular level, the level of mRNA in the duodenal mucosa from the subject since precise molecular analyses have not been performed. was not reduced. Thus, in three subjects with type I xanthi- We recently cloned rat (6) and subsequently human (7) nuria, the primary genetic defects were confirmed to be in xanthine dehydrogenase cDNA and determined the primary the xanthine dehydrogenase gene. (J. Clin. Invest. 1997. 99: structures. The human gene was mapped to chromosome 2p22 2391–2397.) Key words: urolithiasis • aldehyde oxidase • or 2p23 (8). Human aldehyde oxidase cDNA was also cloned sulfite oxidase • molybdenum cofactor • hypouricemia and sequenced, although it was initially claimed as cDNA of xanthine dehydrogenase (9) and was subsequently found to be Introduction of aldehyde oxidase (10). As a first step to determine which gene is responsible for each xanthinuria type, we attempted to Human xanthine dehydrogenase (EC 1.1.1.204) catalyzes the define mutations in the human xanthine dehydrogenase gene terminal two steps of the purine degradation pathway; i.e., the in patients with classical xanthinuria. formation of xanthine from hypoxanthine and uric acid from xanthine. Under certain conditions, xanthine dehydrogenase is Methods converted to the oxidase form, known as xanthine oxidase (EC 1.2.3.2). This enzyme has been a focus of extensive studies Subjects used in this study. Subjects 1 and 2 were brothers, 34 and 35 since (a) it is the target of action of a widely used antihyperuri- yr old, respectively, when examined. Their parents were first cousins. cemic drug, allopurinol; (b) it may be responsible for the pro- Subject 1 was first found to be hypouricemic during a routine health duction of superoxide radicals, which are putative pathological examination; he also had mild gastritis and a duodenal ulcer. Subject compounds in various disorders in humans; (c) inherited en- 2 was asymptomatic. Details on subject 3, a 65-yr-old male, and sub- zyme deficiencies have been described in humans; and (d) the ject 4, a 37-yr-old male, have been reported previously in the Japa- nese literature (11, 12). The chronic renal failure of subject 3 had been attributed to diabetes mellitus (11). Subject 4 had a single kid- ney. The parents of subject 4 were not consanguineous, but their fam- Address correspondence to Kimiyoshi Ichida, Second Department of ilies had lived in the neighborhood for generations. As far as we Medicine, The Jikei University School of Medicine, Nishi-Shinbashi, traced, there was no familial relationship between subjects 3 and 4 Minato-ku, Tokyo 105, Japan. Phone: 81-3-3433-1111; FAX: 81-3- and the parents of subjects 1 and 2. Duodenal mucosal tissues were 3433-4297. obtained from subjects 1 and 4 and a control subject during endo- Received for publication 11 November 1996 and accepted in re- scopic examinations. The mucosal tissues were frozen at Ϫ80ЊC im- vised form 28 February 1997. mediately after collection. The control subject from whom the duodenal mucosa sample and the peripheral blood cells were obtained was J. Clin. Invest. a healthy 36-yr-old male. Peripheral blood cells from five other © The American Society for Clinical Investigation, Inc. healthy volunteers were used for isolating DNA as control samples. 0021-9738/97/05/2391/07 $2.00 Liver tissue from a 58-yr-old male tumor patient was used for confir- Volume 99, Number 10, May 1997, 2391–2397 mation of the reactivity of the antiserum against rat xanthine dehy-

Mutation in Xanthine Dehydrogenase 2391

Table I. Primers Used for this Study

Forward Reverse

Name Sequence Position* Name Sequence Position*

Comp20 GTGACAATGACAGCAGACAAACAAGTTT- Comp21 GTGGCTGACTGAGTGGTCATTTGATT- CGTGAGCTGATTG CTGGACCATGGC 20 GCGAATTCGTGACAATGACAGCAGACAAA Ϫ6–15 21 AACTCGAGGTGGCTGACTGAGTGGTC 573–556 64 GTCTCTTAGGAGTGAGG Ϫ49–Ϫ33 65 GTGGCTGACTGAGTGGTC 573–556 22 AACTCGAGGATGGTGGATGCTGTGGA 496–513 13 CACTGGCCATGAACACG 1105–1089 P4 GCGCTGGTTTGCTGGGAAGC 1002–1021 6 GATGCAGCTCCTCTGCC 1486–1470 17 TCAGCCCTCAAGACCAC 1393–1409 27 GCCTCGAGGATATGCCCAACACAAGT 2001–1984 49 CCCAAGGGTCAGTCTGAG 1693–1710 50 AATCCGGTTTGCTGGAAC 2364–2347 28 CAGAATTCGCGAAGGATAAGGTTACT 1968–1986 5 TCCAGCATGCATCGCAC 2486–2470 1 GAGCACTTCTACCTGGA 2221–2237 10 TAGCAGTTGTCCATGTG 2669–2653 30 CGGAATTCGCCCTGGCTGCATATAAG 2440–2457 31 TTCTCGAGCTTGTCAACCTCACTCTT 2961–2944 55 GAAGTTGCAGTGACCTGT 2788–2805 33 TTCTCGAGGTTAGTCTCAAAGCTGTA 3438–3421 34 TGGAATTCTGCCACTGGGTTTTATAG 3385–3404 35 AGCTCGAGCTGCACGGATGG 3852–3833 42 TTTGTCCAGGGCCTTGG 3598–3614 66 AAGACTCTGCTGAGGAC 4026–4010

Primers Comp20, Comp21, 20, and 21 were used for competitive RT-PCR. *Nucleotide residue numbers are according to Ichida et al. (7).

drogenase to human enzyme after hepatectomy. Informed consents Quantification of mRNA for xanthine dehydrogenase using com- for various procedures were obtained from all subjects at the outset petitive reverse transcriptase-PCR. Total RNA was isolated from of the study. duodenal mucosae or B lymphoblasts using Isogen (Nippon Gene, Determination of compounds in serum and urine. Concentrations Tokyo, Japan) according to the manufacturer’s recommendations. of hypoxanthine, xanthine, and uric acid in both serum and urine, and Xanthine dehydrogenase mRNA in the samples was quantified by concentrations of oxypurinol in serum were determined by the HPLC competitive reverse transcriptase (RT)1-PCR using a MIMIC con- method, as described earlier (13). struction kit (Clontech, Palo Alto, CA). The competitive template Xanthine dehydrogenase/oxidase assay. Xanthine dehydrogenase/ was constructed from 2 ng of Bam HI/Eco RI fragment of v-erbB us- oxidase activity was determined by the HPLC method (13) in the lab- ing primers Comp20 and Comp21 (Table I) in the first 16-cycle PCR oratory of Dr. Toshihiro Nishina in Toranomon Hospital. at 94ЊC, 45 s; 60ЊC, 45 s; and 72ЊC, 90 s. After the RT reaction using Allopurinol loading test. Three allopurinol tablets (100 mg each) RNA from the mucosae, the competitive PCR was performed. Thus, were administered after overnight fasting to each subject and the con- the solution containing cDNA of xanthine dehydrogenase synthe- centration of oxypurinol was determined in the serum from a blood sized by the RT reaction was mixed with various concentrations of sample obtained 1 h after the administration of the drug. the competitive template, and 18 cycles of PCR reactions were per- Establishment of B lymphoblast cell lines. Genomic DNA was iso- formed using primers 20 and 21 (Table I). lated from peripheral blood cells using a QIAGEN blood and cell Direct sequencing of cDNA. The primers used for RT-PCR am- culture DNA kit (Qiagen, Hilden, Germany). B lymphoblast cell plification of cDNA of human xanthine dehydrogenase are shown in lines were established from subjects 3 and 4 using Epstein-Barr virus Table I. cDNA was reverse transcribed from total RNA from the transformation. duodenal mucosae or B lymphoblasts and subjected to the PCR reac- Western blotting. The frozen duodenal mucosal tissues were tions. The conditions for PCR were essentially the same as previously thawed before the experiments and homogenized in 5 vol of 100 mM described (15) with an annealing temperature of 62ЊC. The PCR potassium phosphate buffer, pH 7.4, containing 1 mM PMSF (Wako, products were separated by agarose gel electrophoresis, cut out from Osaka, Japan), 0.02 mg/ml each of aprotinin (Miles Inc., Kankakee, the gel, and isolated using GeneClean kit (BIO 101, La Jolla, CA). IL), chymostatin (Peptide Laboratory, Osaka, Japan), pepstatin (Pep- Direct sequencing was performed using a Dye Terminator Cycle Se- tide Laboratory) and antipain (Peptide Laboratory), and 0.1 mg/ml of quencing Kit and an automated sequencer (ABI Imaging Inc., Ana- leupeptin (Peptide Laboratory), by a Potter-Elvehjem homogenizer heim, CA). at 4ЊC. The homogenates were centrifuged for 1 h at 100,000 g. The Direct sequencing of DNA. The amplification of genomic DNA supernatants (20 ␮g) were subjected to gradient SDS polyacrylamide for the analysis of the nucleotide change at nucleotide 2567 was per- gel electrophoresis in which the concentration of polyacrylamide was formed by PCR using the following two primers: 30 (sense primer, varied from 4 to 20%. After electrophoresis, the protein was trans- Table I) and ACTCTGAGAGAGATCCT (antisense primer). These kbp sequence of genomic DNA including the-2 ف blotted from the gel onto a membrane (Clear Blot Membrane-P; primers amplify an Atto, Tokyo, Japan) using a semidry transblot apparatus (Atto). The point of deletion mutation at nucleotide 2567 found in subject 4. The bands for xanthine dehydrogenase were made visible using rabbit an- nucleotide numbers in this paper always indicate those in cDNA but tiserum raised against rat xanthine dehydrogenase and a Vectastain not in genomic DNA. When those numbers are used for genomic ABC alkaline phosphatase kit (Vector Laboratories, Inc., Burlin- DNA, they indicate the nucleotide positions corresponding to such game, CA), essentially according to the supplier’s protocol. As the nucleotide positions in cDNA sequences. The amplified genomic antiserum against rat enzyme was known to react with rat enzyme to DNA was sequenced by the direct sequencing method. form precipitate (14), the same antiserum was confirmed to react also with human enzyme by determination of enzyme activity of the su- pernatant obtained by addition of the antiserum to the human liver 1. Abbreviations used in this paper: SSCP, single-stranded conforma- extract followed by centrifugation (data not shown). tion polymorphism; RT, reverse transcriptase.

2392 Ichida et al.

The amplification of genomic DNA for the analysis of the nucle- whether he should be classified as type I or II since there was otide substitution at nucleotide 3449 was performed by PCR using no allopurinol loading test. Unfortunately, we were unable to the following two primers: AACACCCAATCTGGGCTACA (sense contact him directly, so the classification of his xanthinuria is primer) and CTTATGATCTCCTGTTAGGC (antisense primer). not possible. The renal function of subject 3 was severely im- These primers amplify a 115-bp sequence of genomic DNA including paired, which was attributed to diabetes mellitus but not to the base substitution at nucleotide 3449. Whether the nucleotide at xanthinuria (11). Although data on the allopurinol loading test position 3449 was C or G could be determined by both the direct sequencing and the digestion of the amplified genomic DNA with Aci I. on subject 4 were not included in the previous paper (12), we When the nucleotide position at 3449 is G, the amplified DNA should performed the test on the subject and found that he had type I be digested with Aci I into two fragments, while when it is C, the xanthinuria (Table II). DNA should be resistant to the digestion. Detection of xanthine dehydrogenase protein in duodenal Single-stranded conformation polymorphism analysis after PCR mucosa. The duodenal mucosa samples (from subject 1 and amplification of genomic DNA. For the detection of the nucleotide the control subject) were submitted to Western blot analysis, change at the position of 682, a part of the genomic xanthine dehy- as described in Methods. A clear band corresponding to hu- drogenase gene was amplified by PCR using the following two prim- man xanthine dehydrogenase was detected at 150 kD for the ers: AGACACTCCTCGGAAGCAGC (sense primer) and CGTGT- control sample while no such band was observed for the sam- TCCCCACGACCAGCT (antisense primer). These primers amplify ple from subject 1 (Fig. 1). The bands at 127 and 50 kD in Fig. a 127-bp sequence of xanthine dehydrogenase genomic DNA, including the nonsense base substitution found in subject 1. Then the ampli- 1 were considered to be nonspecific bands because neither the fied DNA was subjected to the single-stranded conformation poly- reported size of the native form of xanthine dehydrogenase morphism (SSCP) analysis according to the previous methods, except nor those of the proteolytic products were 127 or 50 kD (17). for the temperature during the electrophoresis. In the present experi- However, a possibility that either of these bands corresponds ment, the electrophoresis was performed at 37ЊC. After the electro- to the truncated mutational peptide cannot be completely ex- phoresis, the gel was stained with silver as described previously (16). cluded. Quantitation of mRNA for xanthine dehydrogenase. Total Results RNA was separated from the duodenal mucosa samples, after which the RNA samples were subjected to the competitive Clinical tests. The basis of the hypouricemia in subjects 1 and 2 RT-PCR procedure, as described in Methods. While the prim- was sought. Both were found to excrete excessive amounts of ers 20 and 21 should amplify a 579-bp fragment of human xan- xanthine and hypoxanthine (Table II) and their conditions thine dehydrogenase cDNA, the same primers are expected to were diagnosed as xanthinuria. The classification of classical amplify a 441-bp fragment from a competitor template. By xanthinuria is determined by the outcome of an allopurinol comparing the concentrations of the competitor template that loading test (3). Since aldehyde oxidase as well as xanthine de- gave nearly equal fluorescent intensities of the amplified 579- hydrogenase converts allopurinol to oxypurinol, type II but and 441-bp fragments between subject 1 and the control sam- not type I xanthinuria patients lack the ability to produce oxy- ples, we could compare the amounts of mRNA for xanthine purinol from allopurinol (3). Since oxypurinol was detected af- dehydrogenase between the control subject and subject 1. For ter the allopurinol loading test in both subjects 1 and 2 (Table both subject 1 and the control, the fluorescent intensities for II), they were classified as classical xanthinuria type I. 579- and 441-bp fragments were nearly equal at a competitor Although precise clinical data on subjects 3 and 4 have al- fragment concentration of 5.0 ϫ 10Ϫ3 amol competitor/tube ready been published (11, 12), we include some of them in Ta- (Fig. 2). Virtually identical results were obtained in separate ble II since the previous papers were in Japanese. Table II in- experiments where different amounts of total RNA or differ- dicates that subject 3 had xanthinuria, but it is not clear ent PCR cycles were used (data not shown). These data indi-

Table II. Clinical Data from Four Xanthinuria Subjects and a Control Subject

Subject 1 2 3* 4‡ Control

Sex Male Male Male Male Male Age (yr) 34 35 65 37 36 Creatinine clearance (ml/min) 129 138 8.6 1.3 121 Serum urate (mg/dl) 0.1 0.4 0.1 0.1 6.8 Urate clearance (ml/min) 1.13 0.17 NM 0.06 10.2 Uric acid/creatinine in urine (␮mol/mmol) 1 0.6 4 0.2 382 Serum hypoxanthine (␮M) 5.9 7.3 0.8 1.6 6.2 Serum xanthine (␮M) 23.6 15.8 85 38.5 2.7 Hypoxanthine/creatinine in urine (␮mol/mmol) 89 57.8 1.5 38.9 4.2 Xanthine/creatinine in urine (␮mol/mmol) 110 85.7 132 108.2 3.6 Xanthine oxidase activity (nmol/h per mg protein) ND ND ND ND 52 Serum oxypurinol§ (␮g/ml) 2.8 3.1 NM 0.61 2.3 Type of xanthinuria I I Unknown I —

ND, not detected; NM, not measured. *Clinical data for subject 3 were from reference 11. ‡Clinical data for subject 4 were from reference 12. §Serum oxypurinol was measured at 1 h after loading.

Mutation in Xanthine Dehydrogenase 2393

Figure 1. Western blot cDNA for xanthine dehydrogenase was also synthesized by for the detection of xan- RT-PCR, using total RNA obtained from B lymphoblast cell thine dehydrogenase lines from subject 3. The amplification of cDNA from B cell protein. Extract from lines was possible only during the early phase after the estab- the duodenal mucosa lishment of the cell lines. These data probably reflect the tran- obtained from subject 1 sient synthesis of minimal amounts of mRNA for xanthine de- (lane 1) or the control hydrogenase in this type of cell. Sequencing of cDNA from subject (lane 2) was subject 3 verified the fact that he also carried the same non- subjected to SDS polyacrylamide (4–20%) sense mutation CGA (Arg) to TGA (Ter) at codon 228. Re- gel electrophoresis; petitive sequencing experiments showed that his cDNA had each lane contained only the mutational but not normal sequence at this nucleotide 20 ␮g protein. Thereaf- position. On the other hand, cDNA from subject 4 did not ter, the protein was carry the mutation at nucleotide 682 but had a deletion of C at transferred to a mem- nucleotide 2567 (data not shown). This single nucleotide dele- brane and immunostained with antiserum against rat xanthine dehy- tion should cause a frame shift and generate a termination drogenase. Positions of molecular markers are shown at the left side codon from nucleotide 2783. The G to C change at 3449 was of the panel. 150-kD bands correspond to human xanthine dehydro- also present in the sequence of cDNA from subject 4 (data not genase, while other bands are nonspecific. shown). SSCP analysis after PCR amplification of genomic DNA. Genomic DNA from subjects 1–4, the father and the mother of cate that the xanthine dehydrogenase gene is indeed tran- subjects 1 and 2, and two control subjects was submitted to the scribed and retained in the cells of subject 1. We subsequently PCR procedure followed by the SSCP analysis as described in performed an equivalent analysis on the duodenal mucosa Methods. When the electrophoresis was performed at 20ЊC, sample from subject 4. The amount of mRNA in the mucosa the separation of the mutant and normal sequence was not from subject 4 was not significantly different from the sample successful. However, when the temperature was raised to from the control subject (data not shown). 37ЊC, the mutant sequence was clearly separated from the nor- Direct sequencing of cDNA. Since mRNA was confirmed mal sequence. As in Fig. 3, subjects 1 and 2 were shown to be to be present in the cells of subjects 1 and 4, we amplified and homozygous concerning the nonsense mutation at codon 228, sequenced all the coding regions of the xanthine dehydroge- while both the father and the mother of subjects 1 and 2 were nase gene by RT-PCR. The 11 primer pairs used for the ampli- heterozygous. Although the sequencing of cDNA from subject fication are shown in Table I. Within the entire coding region, 3 showed that it contained only the mutant sequence, the ge- only two base changes were identified for subject 1 when com- nomic DNA from the same subject indicated that he was in pared with the sequence reported previously (7). Thus, a C to fact heterozygous as to this mutation. The genomic DNA from T base change at nucleotide position 682 and a G to C change subject 4 had no alleles with the nonsense mutation at codon at 3449 (data not shown) were found (7). The former nucle- 228 (Fig. 3). otide substitution should generate a nonsense substitution Analysis of genomic DNA at nucleotide 2567. Genomic from CGA (Arg) to TGA (Ter) at codon 228, while the latter DNA from subject 4 was amplified as described in Methods should cause a missense base change from CGC (Arg) to CCC for the analysis of the single nucleotide deletion at nucleotide (Pro) at codon 1150. The latter base replacement was also ob- 2567. The direct sequencing of the amplified genomic DNA re- served in the sequence of human xanthine dehydrogenase re- vealed that subject 4 had the deletion of C at nucleotide 2567 cently reported (18). as a homozygous state (data not shown).

Figure 2. Quantification of xanthine dehydrogenase mRNA. 0.5 ␮g of total RNA from the duodenal mucosa of the control subject (lanes 1–6) or subject 1 (lanes 7–12) was used as test templates for cDNA synthesis. Procedures for the competitive RT-PCR were as described in Methods. The final PCR products were subjected to 1% agarose gel electrophoresis and stained with ethidium bromide. The amount of the competitor template added per tube was 0.2 amol (lanes 1 and 7), 2.0 ϫ 10Ϫ2 amol (lanes 2 and 8), 1.0 ϫ 10Ϫ2 amol (lanes 3 and 9), 5.0 ϫ 10Ϫ3 amol (lanes 4 and 10), 2.5 ϫ 10Ϫ3 amol (lanes 5 and 11), or 1.25 ϫ 10Ϫ3 amol (lanes 6 and 12). XDH, xanthine dehydrogenase.

2394 Ichida et al.

zyme. Results also showed that none of the 10 subjects, including the 4 subjects with xanthinuria, had G at nucleotide 3449 because DNA from all the individuals was resistant to the cleavage (Fig. 4). Therefore, G at nucleotide 3449 seems to be either a rare polymorphism, a nucleotide change generated during the establishment of the cDNA library, or a sequencing error.

Discussion

In the present investigation, we identified a C to T base substitution at nucleotide 682 (nucleotide number in cDNA) and a deletion of C at nucleotide 2567. The former mutation should cause a CGA (Arg) to TGA (Ter) nonsense mutation at codon 228, while the latter mutation should cause a frameshift from codon 856, and a termination codon is encountered at codon 928. Human xanthine dehydrogenase is a large molecule con- sisting of 1333 amino acids (7). Mapping of the functions on xanthine dehydrogenase was performed for three peptide do-

mains generated by the protein cleavage (2). The NH2-termi- Figure 3. Genomic DNA was extracted from either peripheral blood nal 20-kD domain includes a 2Fe/2S nonheme iron binding mononuclear cells or B cell lines from subject 1 (S1), subject 2 (S2), site, while the adjacent 40-kD and the COOH-terminal 85-kD the father (F) and the mother (M) of subjects 1 and 2, subject 3 (S3), domains include flavin binding and molybdenum cofactor subject 4 (S4), and two control subjects (C). The genomic DNA was binding domains, respectively (2). Recently, it was reported submitted to the PCR-SSCP procedure and the gel was stained with that many segments in xanthine dehydrogenase protein con- silver as described in Methods. Mutant specific (M) and normal spe- tact molybdenum cofactor (19). Both nonsense and deletion cific (N) bands are indicated with arrows. Note that a single double mutations found here may cause the cells to synthesize the strand DNA fragment should generate two bands in this system, one truncated peptides. However, even if they are synthesized, corresponding to the sense and the other to the antisense sequence. normal functions are not likely to be retained. Flavin and molybdenum cofactor binding sites would be missing in the xanthine dehydrogenase peptide of subject 1 and the peptide of Analysis of genomic DNA at nucleotide 3449. Genomic subject 4 would lack molybdenum cofactor binding sites. DNA from subjects 1–4 and six control subjects was amplified The results of the Western blot and quantitative RT-PCR as described in Methods for the analysis of the base substitu- analyses were consistent with the mutation we identified. Since tion at nucleotide 3449. Direct sequencing of the amplified ge- subjects 1 and 4 were homozygous for the nonsense and dele- nomic DNA showed that all 10 subjects were homozygous for tion mutations, respectively, all mRNA synthesized in each C at nucleotide 3449 (data not shown). In addition, whether subject should possess a single defect. The nonsense mutation the nucleotide position at 3449 was G could be determined by in mRNA would cause a failure in synthesizing mature protein the digestion of the amplified DNA with Aci I restriction en- detectable by the Western blot method. Although some re-

Figure 4. Analysis of genomic DNA at nucleotide 3449. Genomic DNA was extracted from subjects 1–4 (lanes 1–4) and six control subjects (lanes 5–10). Thereafter, a part of the sequence including the nucleotide 3449 was amplified as described in Methods. The amplified DNA was digested with Aci I under the conditions recommended by the supplier and submitted to agarose gel electrophoresis. The arrow indi- cates the bands at 115 bp corresponding to the uncleaved xanthine dehydrogenase DNA segment.

Mutation in Xanthine Dehydrogenase 2395

ports described decreases in the amounts of mRNA by non- xanthine dehydrogenase gene contains CCAAT/enhancer sense mutations (20), they probably reflected the relative in- binding protein (C/EBP), IL-6, and nuclear factor–␬B binding stability of mRNA. It is not surprising that such decreases in sites, consensus sequences related to inflammation and acute the amounts of mRNA are not found in other cases with non- phase responses (31). When such physiological and pathologi- sense and deletion mutations. Therefore, our data regarding cal roles of xanthine dehydrogenase/oxidase are considered, the enzyme protein and mRNA are essentially what would be the presence of individuals who lack the enzyme is valuable. expected from the nonsense and deletion mutations we found. When the above hypotheses are evaluated, one should keep in By the analysis of the genomic DNA, the genotypes of the mind that there are individuals in which no generation of ac- four subjects were determined. Both subjects 1 and 2 were ho- tive oxygen species through this pathway is present. Our mozygous for the nonsense mutation at codon 228, while sub- present study confirms that in such individuals, the ability to ject 3 was a compound heterozygote. One of the alleles in sub- synthesize xanthine dehydrogenase is indeed genetically im- ject 3 had the nonsense mutation while the other mutation paired. remains to be identified. Since the codon 228 nonsense mutation was exclusively observed in the cDNA from his B cell line, Acknowledgments the analysis of the entire genomic DNA is likely to be required to identify the other mutation. Subject 4 was found to be ho- We thank Dr. Shigeaki Muto (Department of Nephrology, Jichi Med- mozygous for the deletion mutation in both of the alleles. ical School, Tokyo, Japan) for supplying a blood sample from subject Since subjects 1, 2, and 4 had classical xanthinuria type I (Ta- 3, and Drs. Ryozo Sakuma and Toshihiro Nishina (Toranomon Hos- ble I), the genetic defects in the xanthine dehydrogenase gene pital, Tokyo, Japan) for performing the xanthine dehydrogenase/oxi- were confirmed to cause classical xanthinuria type I. dase assays. We also thank Dr. M.W. Schein (Rockville, MD) for re- viewing this manuscript. Although it is not surprising that we identified the same mutation in subjects 1 and 2, the presence of the same mutation in subject 3 is of interest. Codon 228 in the human xan- References thine dehydrogenase gene may be a hot spot for mutation 1. Dent, C.E., and G.R. Philport. 1954. Xanthinuria, an inborn error (or de- since C in the CpG duplet is often methylated in the germline, viation) of metabolism. Lancet. i:182–185. and the deamination of methyl cytosine would convert the 2. Simmonds, H.A., S. Reiter, and T. Nishino. 1995. Hereditary xanthinuria. base to thymine (21). Therefore, the same mutation found in In The Metabolic Basis of Inherited Disease. 7th ed. C.R. Scriver, A.L. Beau- det, W.S. Sly, and D. Valle, editors. McGraw-Hill, Inc., New York. 1781–1797. the family of subjects 1 and 2 and in subject 3 may reflect re- 3. Reiter, S., H.A. Simmonds, N. Zollner, S.L. Braun, and M. Knedel. 1990. current germline mutations at the mutational hot spot. Alter- Demonstration of a combined deficiency of xanthine oxidase and aldehyde oxi- natively, the nonsense base substitution found in the two sepa- dase in xanthinuric patients not forming oxypurinol. Clin. Chim. Acta. 187:221–234. 4. Jonson, J.L., and S.K. Wadman. 1995. Molybdenum cofactor deficiency rate families may have originated from a single mutation. The and isolated sulfite oxidase deficiency. In The Metabolic Basis of Inherited Dis- latter possibility cannot be excluded since, in some autosomal ease. 7th ed. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle, editors. recessive diseases, the same mutation found in separate fami- McGraw-Hill, Inc., New York. 2271–2283. 5. Harkness, R.A., G.M. McCreanor, D. Simpson, and I.R. MacFadyen. lies in an ethnic group were derived from a single ancestor 1986. Pregnancy in and incidence of xanthine oxidase deficiency. J. Inherited gene (22–24). Metab. Dis. 9:407–408. Identification of mutations in the xanthine dehydrogenase 6. Amaya, Y., K. Yamazaki, M. Sato, K. Noda, T. Nishino, and T. Nishino. 1990. Proteolytic conversion of xanthine dehydrogenase from the NAD-depen- gene in classical xanthinuria type I provides a first molecular dent type to the O2-dependent type. J. Biol. Chem. 265:14170–14175. basis to the complicated classification of xanthinuria. Thus, al- 7. Ichida, K., Y. Amaya, K. Noda, S. Minoshima, T. Hosoya, O. Sakai, N. though the primary causes of classical xanthinuria type II and Shimizu, and T. Nishino. 1993. Cloning of the cDNA encoding human xanthine dehydrogenase (oxidase): structural analysis of the protein and chromosomal molybdenum cofactor deficiency are still to be investigated, location of the gene. Gene. 133:279–284. the cause of classical xanthinuria type I was confirmed to be lo- 8. Minoshima, S., Y. Wang, K. Ichida, N. Shimizu, T. Nishino, and N. cated in the xanthine dehydrogenase gene. It is of interest to Shimizu. 1995. Mapping of the gene for human xanthine dehydrogenase (oxidase) (XDH) to the band p23 of chromosome 2. Cytogenet. Cell Genet. 68:52–53. note that defects in the sulfuration of desulfo molybdenum hy- 9. Wright, R.M., G.M. Vaitaitis, C.M. Wilson, T.B. Repine, L.S. Terada, droxylases have been hypothesized as the underlying mecha- and J.E. Repine. 1993. cDNA cloning, characterization, and tissue-specific ex- nism in type II xanthinuria, since a similar defect is present in pression of human xanthine dehydrogenase/xanthine oxidase. Proc. Natl. Acad. Sci. USA. 90:10690–10694. Drosophila melanogaster (25). In mammals, however, it is still 10. Turner, N.A., W.A. Doyle, A.M. Ventom, and R.C. Bray. 1995. Proper- unclear how a molybdenum cofactor of xanthine dehydroge- ties of rabbit liver aldehyde oxidase and the relationship of the enzyme to xan- nase and aldehyde oxidase is converted from the desulfo form thine oxidase and dehydrogenase. Eur. J. Biochem. 232:646–657. 11. Oono, S., S. Muto, T. Nishina, R. Sakuma, H. Yamanaka, K. Nishioka, to the sulfo form, or how the cofactor is introduced into the K. Tabei, and Y. Asano. 1992. A case of hereditary xanthinuria with chronic re- apo enzymes. Further molecular studies on individuals with nal failure due to diabetic nephropathy. Intern. Med. (Tokyo). 70:192–195. classical xanthinuria type II and molybdenum cofactor defi- 12. Kono, S., T. Hosoya, K. Kodama, A. Matsumoto, Y. Oda, Y. Ogura, S. Sakai, and T. Miyahara. 1984. Two cases of xanthinuria in brothers. Naikagak- ciency are likely to contribute to a more reliable classification kaizasshi. 73:33–38. of hereditary xanthinuria and the further understanding of the 13. Kojima, T., T. Nishina, M. Kitamura, T. Hosoya, and K. Nishioka. 1984. biochemical relationship between those enzymes. Biochemical studies on the purine metabolism of four cases with hereditary xanthinuria. Clin. Chim. Acta. 137:189–198. Xanthine dehydrogenase is known to be converted to the 14. Ichikawa, M., T. Nishino, T. Nishino, and A. Ichikawa. 1992. Subcellular oxidase form by the proteolytic cleavage and the oxidation of localization of xanthine oxidase in rat hepatocytes: high-resolution immuno- cystein residues (6). The oxidase form has been implicated as electron microscopic study combined with biochemical analysis. J. Histochem. Cytochem. 40:1097–1103. the main generator of the active oxygen species, which are 15. Saiki, R.K., D.H. Gelfand, S. Stoffel, S.J. Sharf, R. Higuchi, G.T. Horn, considered to be pathogenic in such conditions as postischemic K.B. Mullis, and H.A. Erlich. 1988. Primer-direct enzymatic amplification of reperfusion tissue injury (26, 27), adult respiratory distress syn- DNA with a thermostable DNA polymerase. Science (Wash. DC). 239:487–491. 16. Terai, C., M. Hakoda, H. Yamanaka, N. Kamatani, M. Okai, F. Taka- drome (28), and lung injury resulting from influenza virus in- hashi, and S. Kashiwazaki. 1995. Adenine phosphoribosyltransferase deficiency fection (29, 30). Moreover, the 5Ј flanking region of the human identified by urinary sediment analysis: cellular and molecular confirmation.

2396 Ichida et al. Clin. Genet. 48:246–250. The Metabolic Basis of Inherited Disease. 7th ed. C.R. Scriver, A.L. Beaudet, 17. Hille, R., and T. Nishino. 1995. Xanthine oxidase and xanthine dehydro- W.S. Sly, and D. Valle, editors. McGraw-Hill Inc., New York. 1707–1724. genase. FASEB J. 9:995–1003. 25. Wahl, R.C., C.K. Warner, V. Finnerty, and K.V. Rajagopalan. 1982. 18. Saksela, M., and K.O. Raivio. 1996. Cloning and expression in vitro of Drosophila melanogaster ma-1 mutants are defective in the sulfuration of human xanthine dehydrogenase/oxidase. Biochem. J. 315:235–239. desulfo Mo hydroxylases. J. Biol. Chem. 257:3958–3962. 19. Romao, M.J., M. Archer, I. Moura, J.J.G. Moura, J. LeGall, R. Engh, 26. MacGowan, S.W., M.C. Regan, C. Malone, O. Sharkey, L. Young, T.F. M. Schneider, P. Hof, and R. Huber. 1995. Crystal structure of the xanthine ox- Gorey, and A.E. Wood. 1995. Superoxide radical and xanthine oxidoreductase idase–related aldehyde oxido-reductase from D. gigas. Science (Wash. DC). activity in the human heart during cardiac operations. Ann. Thorac. Surg. 60: 270:1170–1176. 1289–1293. 20. Brawerman, G. 1989. mRNA decay: finding the right targets. Cell. 57:9–10. 27. McCord, J.M. 1985. Oxygen-derived free radicals in post-ischemic tissue 21. Coulondre, C., J.H. Miller, P.J. Farabaugh, and W. Gilbert. 1978. Mo- injury. N. Engl. J. Med. 312:159–163. lecular basis of base substitution hotspots in Escherichia coli. Nature (Lond.). 28. Grum, C.M., R.A. Ragsdale, L.H. Ketai, and R.H. Simon. 1987. Plasma 274:775–780. xanthine oxidase activity in patients with ARDS. J. Crit. Care. 2:22–26. 22. Kamatani, N., M. Hakoda, S. Otsuka, H. Yoshikawa, and S. Kashi- 29. Oda, T., T. Akaike, T. Hamamoto, F. Suzuki, T. Hirano, and H. Maeda. wazaki. 1992. Only three mutations account for almost all defective alleles caus- 1989. Oxygen radicals in influenza-induced pathogenesis and treatment with ing adenine phosphoribosyltransferase deficiency in Japanese patients. J. Clin. pyran polymer-conjugated SOD. Science (Wash. DC). 244:974–976. Invest. 90:131–136. 30. Akaike, T., M. Ando, T. Oda, T. Doi, S. Ijiri, S. Araki, and H. Maeda. 23. Satokata, I., K. Tanaka, and Y. Okada. 1992. Molecular basis of group 1990. Dependence on O2-generation by xanthine oxidase of pathogenesis of in- A xeroderma pigmentosum: a missense mutation and two deletions located in a fluenza virus infection in mice. J. Clin. Invest. 85:739–745. zinc finger consensus sequence of the XPAC gene. Hum. Genet. 88:603–607. 31. Xu, P., T.P. Huecksteadt, and J.R. Hoidal. 1996. Molecular cloning and 24. Simmonds, H.A., A.S. Sahota, and K.J. Van Acker. 1995. Adenine characterization of the human xanthine dehydrogenase gene (XDH). Geno- phosphoribosyltransferase deficiency and 2,8-dihydroxyadenine lithiasis. In mics. 34:173–180.

Mutation in Xanthine Dehydrogenase 2397